It’s one of those things we say: “The Hawaiian Islands are the most remote landmass on earth.” The truth, as usual, is more complicated than that.

First, let’s deal with the simplest facts. The Hawaiian Islands, by virtue of being an archipelago, many parts of which are actually within eyesight of one another, simply cannot include the most remote landmass on earth. That title, it turns out, belongs to Bouvet Island, a tiny spit of rock in the South Atlantic that, curiously enough, belongs to Norway and is 994 miles from its nearest neighbor, Queen Maud Land Antarctica. Of course, no one lives in Queen Maud Land. Bouvet’s nearest inhabited neighbor, the island of Tristan da Cunha, is 1,404 miles away.

Hawaii isn’t even the most remote archipelago. That distinction belongs to the same Tristan da Cunha, which is 1,740 miles from South Africa, and is part of the British Overseas Territory of Saint Helena, Ascencion and Tristan da Cunha. (Napoleon died in exile in Saint Helena, proving this South Atlantic island to be much more remote than Mediterranean Elba.) Tristan da Cunha is also the most remote inhabited island in the world.

Most of this remoteness data comes from a fascinating (but is it accurate?) Wikipedia entry called “Extreme points of Earth”. In this article, you’ll also learn that the South Pacific Oceanic Pole of Inaccessibility–the point in the ocean farthest from any landmass, also known as Point Nemo, in deference to Jules Verne–lies 1,670 miles from Dulcie Island, which is part of the Pitcairn Islands. Pitcairn, because of the story of the Mutiny on the Bounty, has become a kind of metonym for remoteness at sea, much like Timbuktu has come to represent remoteness on land. (Dulcie was briefly occupied by the starving  survivors of the Essex, a whaleship whose sinking by a sperm whale in 1820 served as the inspiration for Melville’s Moby Dick.)

That’s not to say Hawaii’s not remote, of course. Honolulu, for example, is the most remote city in the world of at least 500,000 people. San Francisco, the next city of comparable size, is 2,387 miles away. And, although Easter Island, 2,180 miles off the coast of Chile, is usually considered the island farthest from a continental land mass, Midway, in the Northwest Hawaiian Islands lies 2,500 miles from Tokyo and 3,200 miles from San Francisco. Anyway you look at it, that’s remote.

But the most compelling measure of the remoteness of Hawaii isn’t in miles, but it genes. Hawaii’s isolation has made it the world capital of endemism. For its size, Hawaii has the highest percentage of species that exist nowhere else on Earth. (Sadly, this also makes it the extinction capital of the world; if a species disappears here, it has probably disappeared everywhere.) This high level of endemism is fueled, of course, by the Islands’ remoteness. Few species were able to survive the long float or flight to get here (let alone get here as a mating pair for non-cloning species.) This permitted the rapid speciation that accounts for the wild variation in Hawaiian honeycreepers. But it also limits the number of potential immigrant species.

The genetic remoteness of Hawaii is highlighted in an astonishing paper in Global Ecology and Biogeography. In this paper, Sara Wood, of the University of Bristol, and her colleagues use computer modeling techniques to predict the “global connectivity” of coral species. Basically, they try to account for the likely routes of the colonization of coral reefs around the world. To do this, their model takes into account biological factors, such as the “competency” of coral larvae (coral larvae are only viable for a short window of time,) biogeographic factors (ocean currents and distances limit the potential successful dispersal of coral larvae,) and time frames (coral spawn a limited, albeit enormous, number of times per year.)

Woods computer model of coral distribution is a convenient stand-in for real world remoteness. By her measure, the region surrounding the South Pacific Oceanic Pole of Inaccessibility represents the most genetically remote place on Earth. As she puts it: “The central Pacific was an almost complete barrier to dispersal, only rarely breached westward from the Galapagos to Marquesas Islands.”

According to Woods, though, as far as coral dispersal patterns go, Hawaii is nearly as remote. The great distances and unfavorable currents that separate Hawaii from its nearest neighbors almost completely isolate our local reefs from the potential colonization from elsewhere. Almost.

And here’s the most amazing finding from Wood’s modeling of coral dispersal: Johnston Atoll is “the sole ‘stepping stone’ into Hawaii.” That means, if her study is correct, every species of coral found in Hawaii descends from a species that arrived in the Islands from tiny Johnston Atoll. Were it not for this “stepping stone”, Hawaii might have no coral at all. How’s that for remote?